Bellabeat

Introduction

This is a case study for Google Data Analyst Certificate:

Bellabeat is a high-tech company that manufactures health-focused smart products. It is a successful small company, but they have the potential to become a larger player in the global smart device market. Since it was founded in 2013, Bellabeat has grown rapidly and quickly positioned itself as a tech-driven wellness company for women

Business Task

Our task is to focus on a Bellabeat product and analyze smart device usage data in order to gain insight into how people are already using the smart devices. Then, using this information, we need to recommend how these trends can help Bellabeat marketing strategy.

Prepar data

ROCCC analysis

Reliability : LOW – dataset was collected from 30 individuals whose gender is unknown.

Originality : LOW – third party data collect using Amazon Mechanical Turk.

Comprehensive : MEDIUM – dataset contains multiple fields on daily activity intensity, calories used, daily steps taken, daily sleep time and weight record.

Current : MEDIUM – data is 5 years old but the habit of how people live does not change over a few years

Cited : HIGH – data collector and source is well documented

Package loading

library(plotly)
library(tidyverse) #wrangle data
library(dplyr) #clean data
library(lubridate)  #wrangle date attributes
library(skimr) #get summary data
library(ggplot2) #visualize data
library(cowplot) #grid the plot
library(readr) #save csv
library(tidyr) #for organizing tabular data
library(janitor) #for data examination and cleaning

Importing data

daily_activity <- read.csv("dailyActivity_merged.csv")
sleep_day <- read.csv("sleepDay_merged.csv")
weight <- read.csv("weightLogInfo_merged.csv")
hourly_steps <- read.csv("hourlySteps_merged.csv")

Data Verification

Examining the first few rows of every data set

head(daily_activity)

head(sleep_day)

head(weight)

head(hourly_steps)

Check for Na

sum(is.na(daily_activity))

[1] 0

sum(is.na(sleep_day))

[1] 0

sum(is.na(weight))

[1] 65

sum(is.na(hourly_steps))

[1] 0

Check for duplicate

sleep_day have 3 duplicates

#CHECK FOR DUPLICATES

sum(duplicated(daily_activity))

[1] 0

sum(duplicated(sleep_day))

[1] 3

sum(duplicated(weight))

[1] 0

sum(duplicated(hourly_steps))

[1] 0

daily activity extra 3 users ,sleep day less 6 users, weight less 22 users

#daily activity extra 3 users ,sleep day less 6 users, weight less 22 users
n_distinct(daily_activity$Id)

[1] 33

n_distinct(sleep_day$Id)

[1] 24

n_distinct(weight$Id)

[1] 8

n_distinct(hourly_steps$Id)

[1] 33

Process

Romove duplicates from sleep_day

sleep_day <- sleep_day[!duplicated(sleep_day), ]

#check
sum(duplicated(sleep_day))

[1] 0

Make useful columns and df

Convert Activity Date into date format and add a column for day of the week

Add column for hours in hourly_steps

hourly_steps$hour <- as.POSIXct(hourly_steps$ActivityHour,format=('%h'))

We will classify the average user’s sleep into three patterns

sleep_day_new <- sleep_day %>% group_by(Id) %>% 
  summarise(avg_time_sleep= mean(TotalMinutesAsleep)) %>%
  mutate(Categores=case_when(
    avg_time_sleep < 300 ~ "Unhealthy Sleep",
    avg_time_sleep >= 320 & avg_time_sleep <=  420 ~ "Normal Sleep",
    avg_time_sleep > 420 ~ "Healthy Sleep"))

We will separate observations into fitness groups based on walking lifestyle: “Sedentary, Needs Improvment, Active, Highly Active”.

Steps_categores <- daily_activity %>% group_by(Id) %>% 
  summarise(avg_steps=mean(TotalSteps)) %>% 
  mutate(level_steps= case_when(
    avg_steps < 5000 ~ "Sedentary",
    avg_steps >=5000 & avg_steps < 10000 ~ "Needsimprovment",
    avg_steps >=10000 & avg_steps < 12500 ~ "Active",
    avg_steps >=12500 ~ "Highly active"))

marge data table ()

merged1 <- merge(daily_activity,sleep_day,by = c("Id"),all= TRUE)

data_marged <- merge(merged1,weight,by = c("Id"),all = TRUE)

Convert Activity Date into date format and add a column for day of the week

data_marged <- data_marged %>%  
  mutate( weekday = weekdays(as.Date(ActivityDate,"%m/%d/%Y")))

#Arrange the days of the week to use in chart
data_marged$weekday <- factor(data_marged$weekday, levels= c("Sunday","Monday","Tuesday", "Wednesday", "Thursday", "Friday", "Saturday"))

analyze and chart

Statistics summary mean, median, min, max for Data_marged

data_marged %>% 
  dplyr:: select(TotalDistance,TotalMinutesAsleep,TotalSteps,TotalTimeInBed,Calories,weekday,WeightPounds,BMI) %>%
  summary()

 TotalDistance    TotalMinutesAsleep   TotalSteps    TotalTimeInBed     Calories         weekday    
 Min.   : 0.000   Min.   : 58.0      Min.   :    0   Min.   : 61.0   Min.   :   0   Sunday   :5610  
 1st Qu.: 3.910   1st Qu.:400.0      1st Qu.: 5832   1st Qu.:421.0   1st Qu.:1850   Monday   :5609  
 Median : 6.820   Median :442.0      Median :10199   Median :457.0   Median :2046   Tuesday  :7004  
 Mean   : 6.415   Mean   :433.8      Mean   : 9373   Mean   :458.2   Mean   :2103   Wednesday:6988  
 3rd Qu.: 8.350   3rd Qu.:477.0      3rd Qu.:12109   3rd Qu.:510.0   3rd Qu.:2182   Thursday :6930  
 Max.   :28.030   Max.   :796.0      Max.   :36019   Max.   :961.0   Max.   :4900   Friday   :5632  
                  NA's   :971                        NA's   :971                    Saturday :5616  
  WeightPounds        BMI       
 Min.   :116.0   Min.   :21.45  
 1st Qu.:134.9   1st Qu.:23.89  
 Median :135.6   Median :24.00  
 Mean   :139.6   Mean   :24.42  
 3rd Qu.:136.7   3rd Qu.:24.21  
 Max.   :294.3   Max.   :47.54  
 NA's   :8881    NA's   :8881   

Let’s look at how active users are per hour in total steps. From 5 p.m. to 7 p.m. users take the most steps

ggplot(data= hourly_steps,aes(x=hour, y=StepTotal,fill=hour))+geom_bar(stat = "identity")+ labs(title = "Steps by Hour",x="Hours",Y="Steps")

Let’s look at categories users are per steps .54.5% from users Needs Improvment

plot_ly(Steps_categores,labels= ~level_steps, value=~avg_steps,type = 'pie', textposition = 'outside',textinfo = 'label+percent') %>% 
  layout(title= 'Users categories by steps')

ggplot(data = data_marged, aes(x=weekday,y=TotalSteps))+geom_bar(stat="identity",fill='steelblue')+labs(title= 'Weeklyday Steps',x="Weekday",y="Steps")

Let’s look at categories users are per average sleep hours .

sleep_day_new$Categores <- factor(sleep_day_new$Categores, levels= c("Unhealthy Sleep","Normal Sleep","Healthy Sleep"))

plot_ly(sleep_day_new,labels= ~ Categores, value= ~ avg_time_sleep,type = "pie", textposition = 'outside',textinfo = 'label+percent') %>% 
  layout(title='usere categories by hourly sleep')

The more active you are, the more steps you take, and the more calories you’ll burn. This is an obvious fact, but we can still look at the data to find anything interesting. Here we see that some users have similar weights, but some of them burn more than 2500 calories and nearly 20,000 steps to reach a weight of 60 kg, others only need to burn more than 1500 calories and only about 10,000 steps, and there are those who weigh up to 120 kg and they can burn over 2000 calories with much fewer steps even less than 5000 steps

ggplot(data=data_marged,aes(x=TotalSteps,y=Calories,color=WeightKg))+ geom_point()+stat_smooth(method = lm)+
  scale_color_gradient(low = 'green',high = 'red')+labs(title = "Steps vs Calories",x="Steps",y="Calories")

`geom_smooth()` using formula 'y ~ x'

And about the rumor that a person who sleeps a lot burns fewer calories, we note that there is no significant correlation between sleeping more and burning fewer calories.

ggplot(data=data_marged,aes(x=Calories,y=TotalMinutesAsleep,color=TotalMinutesAsleep))+ geom_point()+stat_smooth(method = lm)+
  scale_color_gradient(low = 'green',high = 'red')+labs(title = "Sleep Vs Calories",X="Calories" ,y="Sleep Minutes")

`geom_smooth()` using formula 'y ~ x'
Warning: Removed 971 rows containing non-finite values (stat_smooth).
Warning: Removed 971 rows containing missing values (geom_point).